I'm not sure this would fall primarily under recommenders... I would assume 
Facebook is doing "look-ahead" on connections.  i.e. A->B, B->C, so suggest 
A->C.  Then they weight the suggestions by the number of indirect links between 
A and C and probably other factors (which is where the generic "recommender" 
stuff may come in).  I think the bigger challenge here is storing the 
connections in such a way that the lookahead is easy, and I don't think Lucene 
will help you much there.

I've always assumed the right approach for these systems is to "blow out" the 
connections in a db, i.e. if I care about three levels, I insert a row for 
direction connections, connections resulting from direct, and connections 
resulting from that.  Storage is cheap, disk speed is low, etc.  But not sure 
if there's a more intelligent way to do that.

It reminds me a bit of another common problem I don't think I've seen an 
efficient system for yet - AJAX prefix matching (I type McD and you search 
millions of entries to bring back McDonalds and others).  People say use NGram 
for that, but there must be some sort of tree like data structure which would 
be more efficient.

-----Original Message-----
From: Glen Newton [mailto:glen.new...@gmail.com] 
Sent: Tuesday, March 17, 2009 9:38 AM
To: java-user@lucene.apache.org
Subject: Re: "People you might know" ( a la Facebook) - *slightly offtopic*

You might try looking in a list that talks about recommender systems.
Google hits:
- http://en.wikipedia.org/wiki/Recommendation_system
- ACM Recommender Systems 2009 http://recsys.acm.org/
- A Guide to Recommender Systems
http://www.readwriteweb.com/archives/recommender_systems.php

2009/3/17 Aaron Schon <aaron_sc...@yahoo.com>:
>
> Hi all, Apologies if this question is off-topic, but I was wondering if there 
> is a way of leveraging Lucene (or other mechanism) to store the information 
> about connections and recommend People you might know as done in FB or LI.
>
> The data is as follows:
>
> john_sm...@somedomain.com, jane_...@otherdomain.com
>
>
> john_sm...@somedomain.com, frank_jo...@someotherplace.com
>
> and so on...
>
> how would I go about recommending Jane Doe connecting to Frank Jones?. Hope 
> you can help a newbie by pointing where I should be looking?
>
> Thanks in advance,
> AS
>
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org
> For additional commands, e-mail: java-user-h...@lucene.apache.org
>
>



-- 

-

---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-user-h...@lucene.apache.org

Reply via email to